Thermomigration in Eutectic Tin-Lead Flip Chip Solder Joints

Transcription

1 Thermomigration in Eutectic Tin-Lead Flip Chip Solder Joints Dan Yang, M. O. Alam, B. Y. Wu and Y. C. Chan* Department of Electronic Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon Tong, Hong Kong * Abstract The thermomigration of eutectic tin-lead flip chip solder joints under ambient temperatures of 20 C, 0 C and 150 C was investigated in terms of microstructural evolution. No significant thermomigration occurred after 0 h at 20 C and at 0 C. Only microstructural coarsening was observed in solder joints for these cases. However, Pb thermomigration and phase segregation were observed in solder joints after only 50 h at 150 C. Measurements showed that the temperature difference between the chip side and the substrate side reached 8.1 C (a temperature gradient of about 2700 C/cm across the solder joints) at an ambient temperature of 150 C. It is believed that Pb atoms migrated from the chip side (the hot side) to the substrate side (the cold side) under a temperature gradient of 2700 C/cm. 1. Introduction With the trend towards higher integration and further miniaturization in the microelectronics industry, the cross sectional area of conductive lines on the chip has been decreased significantly. Due to different electrical resistances and thermal dissipations of individual parts within the flip chip interconnection structure, it is possible that the heat accumulated at the chip side is larger than that at the substrate side. This will inevitably lead to a considerable temperature gradient across the solder joints, which can provide a driving force for atomic diffusion to trigger thermomigration. In our previous studies [1], we established the presence of this temperature gradient by observing the melted solder near the chip side. Other research work concerning Joule heating also verified this by finite element simulation [2]. Recently, hot spots near the entrances of Al traces in the solder joints have been detected by a thermal infrared microscope [3], which provides further support for the existence of a temperature gradient across solder joints. Being a complex mass migration process, thermomigration, of which the driving force comes from the energy transported by the moving atoms and the interactions with the usual heat carriers in the lattice [4], has become a new reliability concern for flip chip solder joints. Roush and Jaspal [5] ever reported both Pb and In moved under a temperature gradient in lead-indium solder alloys. Van Gurp and co-workers [6] observed fast thermomigration in In and In alloy films and found that atoms migrated from the hot to the cold areas. The latest reports about thermomigration in solder joints include [7,8,9]. Basaran and co-workers [7] reported that thermomigration in flip chip solder joints may assist or counter electromigration with microstructural observations and marker measurements, and their simulation predicted a temperature gradient of 1500 C/cm across solder joints. Chen s group [8] also proposed that a temperature gradient of 365 C/cm was one of the possible reasons for anode failure in their experiments. However, the individual contribution of thermomigration to the failure of solder joints was not investigated in these papers. Recently, Tu s group observed obvious thermomigration in tin-lead composite solder joints [9] and wires [], whereas the microstructural evolution such as phase separation in eutectic tin-lead solder joints under thermomigration was not characterized. As thermomigration might play an important role in the solder joints of microelectronic devices, this investigation is intended to study thermomigration phenomena in eutectic tin-lead solder joints under Joule heating in terms of microstructural analysis. 2. Experimental The samples used in this study were flip chip test chip kits, including dummy chips (from Flip Chip International) and laminate boards (PB , from Practical Components Incorporation). The chip size was 6 mm 6 mm with a thickness of mm. The Al traces on the chip had a width of 5 µm and a thickness of about 2 µm. The under bump metallization (UBM) layer was composed of a thin film of Al/Ni/Cu, and the via in the passivation layer for this UBM had a diameter of 2 µm. There were four rows of 12 solder bumps along each side of the chip with a pitch of 457 µm. The solder bump material was eutectic tin-lead, and the bump diameter was about 190 µm. The test substrate was a high temperature FR4 board with a thickness of 0.84 mm. The copper pad connected to the flip chip had a width of about 152 µm and a thickness of 35 µm, which was covered with a thin 2 µm nickel-plated layer. A solder assist layer of organic solderability preservation coating was placed on the bond pad area. Both chips and substrates were daisy-chained for electrical continuity. The flip chip was attached to the substrate using flip chip bonding technology by a KarlSuss FCM Flip-Chip bonder. The bonded samples were then reflowed using a five-zone air convection oven (BTU VIP-70 N). In the temperature profile, the peak temperature was 230 C and the time above liquidous was about 60 s. Figure 1 shows a typical bonded sample prepared for testing. Figure 2 is a sketch of flip chip solder joints used for this study. Two pairs of solder joints (joint 5, 6, 7 and 8) were powered with a current of 1.8 A at different ambient temperatures of 20 C, 0 C and 150 C. The applied constant current of 1.8 A corresponds to an average current density, defined by dividing the applied current by the area of the passivation via, of about A/cm 2. A MC-8 model (Tabai Espec Corp.) chamber and a LHT4/30 type (Carbolite) oven were utilized to realize the stable and X/06/$ IEEE Electronic Packaging Technology Conference

2 uniform ambient temperatures of 0 C and 150 C. The temperature difference between the chip side and the substrate side under the current stressing and different ambient temperatures was inspected with a M.O.L.E. thermal profiler by contact temperature measurement. Two micro-thermocouples, with a temperature measurement resolution of 0.1 C, were directly attached on top of the chip surface and the copper trace surface near the powered joints, as shown in Figure 3. An additional ten minutes was needed for thermal stabilization and equilibrium of test samples after the current was applied at a certain ambient temperature. 6mm Figure 1 Flip chip sample used in this study (The arrowed region shows the area which is given as an expanded sketch in Figure 2). Silicon chip e - Electron flow Figure 2 Sketch of solder joints with four solder joints (joint 5 to 8) under current stressing. thermocouples FR4 substrate 6mm Figure 3 Temperature measurement of chip surface using thermocouples. Only un-powered solder joints such as joint 4 and 9 (see Figure 2) were investigated for thermomigration study. After an experiment, the samples were ground and polished with a great deal of care for cross-sectional observations. These cross sections were examined under an optical microscope and with a scanning electron microscope (SEM, Philips XL40 FEG) separately. In addition, the local compositions were analyzed using energy dispersive X-ray (EDX). e - 3. Results and Discussions A current of 1.8 A was applied to the middle four solder joints (joint 5, 6, 7 and 8) at 20 C, 0 C and 150 C. The local temperatures on top of the chip surface and the copper trace surface near the powered joints were inspected and compared. As shown in Table I, the temperature difference between the two surfaces increased with the ambient temperature in this test structure, and it reached 8.1 C at an ambient temperature of 150 C. It should be mentioned that the temperature measurement was not accurate because of the contact resistance between the thermocouples and the surfaces, and the actual temperature gradient was likely to be larger than the measured value because the top of solder joints was covered by the chip. However, our measurements provide an important reference that the effect of Joule heating was significant for these test structures, and the temperature gradient across the powered solder joints may approach 2700 C/cm (8.1 C/30 µm 2700 C/cm). Moreover, owing to good thermal conductivity of silicon chip, this temperature gradient was formed across the adjacent un-powered solder joints. Also, the electrical resistance of the Al traces, solder joints and copper conductors in this test structure were calculated to be approximately 121 mω, mω and mω, respectively. These values represent their relative contributions to the Joule heating. It is clear that the thin Al trace is the primary heat source because of its higher Joule heating effect. Hence, it is believed that a certain temperature gradient existed in the un-powered solder joints such as joint 4 and 9. Table I Temperature measurement results at different ambient temperatures when four solder joints were powered with 1.8 A current Ambient temperature ( C) Chip temperature ( C) Copper trace temperature ( C) Temperature difference between the chip side and the substrate side ( C) Figure 4 and 4 illustrate the original microstructure of solder joints before the experiments (as-reflowed). Fine scale Pb rich phase particles (light regions) are uniformly dispersed in the Sn rich matrix (dark regions). These two regions are α-pb and β-sn phases, respectively. For samples treated at 150 C, cross sectioning was performed when powered joint pairs failed after 50 h of current stressing. However, for samples at 20 C and at 0 C, no joint pair was found to fail even after 0 h. Cross sections of these samples whose stressing was conducted up to 0 h were also prepared for comparison. Figure 5 shows an SEM image of the cross section of unpowered solder joint after 0 h at 20 C. No thermomigration phenomena could be detected. However, phase coarsening was observed after 0 h as compared with the original microstructure in Figure 4. This can be explained as phase coarsening as a result of the temperature Electronic Packaging Technology Conference

3 rise, of which the thermodynamic basis is a combination of the driving force for grain coarsening and the Thomson- Freundlich solubility relationship [11]. Moreover, phase phase coarsening Pb rich phase Figure 4 SEM micrograph of original microstructure of solder joints (as-reflowed), local magnified micrograph. Figure 6 SEM micrograph of phase coarsening in an unpowered solder joint after 0 h at 0 C (joint 4). Obvious Pb thermomigration was detected at 150 C after 50 h, as illustrated in Figure 7 and 7. The Pb rich phase separated from the Sn rich phase completely and accumulated at the bottom of solder joints. It is believed that Pb migrated to the substrate side (the cold side) under the temperature gradient across the un-powered solder joints since no current was applied to them. This was also supported by the EDX analysis. The result is in reasonable agreement with those of thermomigration in tin-lead composite flip chip solder joints reported by Tu s group [9]. They found that Pb moved to the cold side and Sn to the hot side under an estimated temperature gradient of above 00 C/cm at 150 C. For other un-powered solder joints (from joint 1 to 3 and joint to 12), the thermomigration was not apparent except for a large amount of phase coarsening, as shown in Figure 8 and 8. phase coarsening Figure 5 SEM micrograph of phase coarsening in an unpowered solder joint after 0 h at 20 C (joint 4). separation can be accelerated considerably when the homologous temperatures are high. Therefore, more substantial coarsening of un-powered solder joints was observed at 0 C after 0 h, as shown in Figure 6. However, thermomigration phenomena were still not apparent. This is similar to the study by Chiang s group [12]. They noticed that the thermomigration of un-powered solder joints was not clear despite a microstructural change when tin-lead solder joints were stressed at a current density of A/cm 2 at 0 C. Sn wt% Pb wt% Sn wt% Pb wt% Figure 7 SEM micrograph of thermomigration in an unpowered solder joint after 50 h at 150 C (joint 4, when the temperature gradient was 2700 C/cm), local magnified micrograph Electronic Packaging Technology Conference

4 phase coarsening Figure 8 SEM micrograph of phase coarsening in an unpowered solder joint after 50 h at 150 C (joint 3, when the temperature gradient was not large enough to trigger thermomigration), local magnified micrograph. An appropriate explanation for the above is as follows. At a lower ambient temperature, the diffusion of atoms was not fast enough to cause a significant migration within a short time. More importantly, the temperature differences between the chip side and the substrate side at 20 C and 0 C were small. Likewise, at 150 C, the temperature gradients across the solder joints (from joint 1 to 3 and joint to 12) were not high enough due to the increasing distances from the powered solder joints, i.e., the source of heating. This is why no noticeable thermomigration in those solder joints was found under these conditions. In contrast, at 150 C there existed a larger temperature gradient across the solder joint 4 and 9 which were nearest to the heat source. Since the flow of atoms is from the hot side to the cold side if there is a temperature gradient [5,6,7,9,], it is supposed that Pb and Sn atoms all migrated from the hot side to the cold side under this temperature gradient. However, Pb atoms are the dominant diffusion species in the eutectic tin-lead solder above 120 C [13,14,15]. Therefore, Sn atoms migrated slowly and replenished the vacancies due to the depletion of Pb atoms. Macroscopically, the Pb-rich phase migrated to one side and the Sn-rich phase was pushed towards the opposite side. This gives a reasonable explanation to the fact that the effect of thermomigration was apparently visible in these solder joints at 150 C. On the other hand, it is noticeable that a few fine Pb grains are still dispersed in the tin-matrix near the hot end, and Sn grains are dispersed in the lead-matrix near the cold end in Figure 7. This is due to dual-phase precipitation. According to the tin-lead phase diagram [16], dual-phase precipitation is expected to occur when single-phase Pb or single-phase Sn are cooled down from around 180 C. Now, the contribution of thermomigration to the failure of solder joints should be considered here. We make a simple estimation for the contributions by driving forces of thermomigration and electromigration with reference to [9]. Taking the atom diameter (d) as 3-8 cm [9] and a temperature gradient (g) of 2700 C/cm, the thermal energy change by the driving force of thermomigration is calculated to be: ωtm = 3k T = 3kdg 23 = ( Joule) In contrast, taking an effective charge number (Z*) of [17], a resistivity (ρ) of -8 Ω m [17], and the current density (j) of A/m 2, the energy change by the force of electromigration is estimated to be: ωem = Z * eρjd 19 = ( Joule) ω ω 8 (1) The ratio of ω tm to ω em is then: tm em 3.35 = = (3) Consequently, mass flux due to the thermomigration driving force cannot be ignored completely, and the equation for mass flux is described as follows: J = J + Jσ + J (4) massflux em tm where J em is the flux due to the electron wind force, J σ is the flux due to the hydrostatic stress gradient, and J tm is the flux due to the thermomigration. Furthermore, similar to that in an Al interconnection [18,19], the combined effect of electromigration and thermomigration in flip chip solder joints should be another important reliability concern. 4. Conclusions We present results on thermomigration in eutectic tinlead solder joints at different ambient temperatures of 20 C, 0 C and 150 C. For the case of 20 C, phase coarsening was observed in the un-powered joints after 0 h. However, no thermomigration could be detected. For the case of 0 C, more dramatic microstructural coarsening was observed after 0 h, but thermomigration was still not evident. For the case of 150 C, obvious Pb thermomigration was observed after only 50 h. The driving force of thermomigration mainly comes from the temperature gradient across the un-powered solder joints. Temperature measurements showed the existence of a temperature difference of about 8.1 C between the chip side and the substrate side (temperature gradient of about 2700 C/cm across solder joints) at 150 C. At an ambient temperature of 150 C Pb and Sn atoms obtain sufficient diffusion energy and migrate from the hot side to the cold (2) Electronic Packaging Technology Conference

5 side, and the migration of Pb can be observed because of its faster diffusion rate in eutectic tin-lead solders. In contrast, Pb migration at 20 C and at 0 C could not be observed due to the smaller temperature gradients and lower ambient temperature. The energy change by the driving force of thermomigration is estimated to be about Joule compared to that of electromigration ( Joule). The atomic flux due to thermomigration should be considered in the total mass flux since thermomigration plays an important role in the reliability of solder joints in microelectronic devices. Acknowledgments This research was supported by RGC funded CERG project CityU 16/04E (CityU internal ref ). The authors would like to thank Prof. K.N.Tu, University of California, Los Angeles, for his valuable discussions. Also, we are grateful to the proofreading of this manuscript from Prof. Brian and Mrs. Anne Ralph. References 1. M.O.Alam, B.Y.Wu, Y.C.Chan, K.N.Tu, High electric current density-induced interfacial reactions in micro ball grid array (µbga) solder joints, Acta Materialia, Vol. 54 (2006), pp Y.S.Lai, C.L.Kao, Electrothermal coupling analysis of current crowding and Joule heating in flip-chip package assembly, Microelectronics Reliability, In press, (2006). 3. S.H.Chiu, T.L.Shao, C.Chen, Infrared microscopy of hot spots induced by Joule heating in flip-chip SnAg solder joints under accelerated electromigration, Applied Physics Letters, Vol. 88, 0221 (2006). 4. H.Ye, C.Basaran, D.C.Hopkins, Mechanical degradation of microelectronics solder joints under current stressing, International Journal of Solids and Structures, Vol. 40, No. 26 (2003), pp W.Roush, J.Jaspal, Thermomigration in lead-indium solder, Proceedings of the 32 nd Electronic Components Conference, CA, May 1982, pp G.J.Van Gurp, P.J.De Waard, F.J.Du Chatenier, Thermomigration in indium and indium alloy, Journal of Applied Physics, Vol. 58, No. 2 (1985), pp H.Ye, C.Basaran, D.C.Hopkins, Thermomigration in Pb- Sn solder joints under joule heating during electric current stressing, Applied Physics Letters, Vol. 82 (2003), pp T.L.Shao, Y.H.Chen, S.H.Chiu, C.Chen, Electromigration failure mechanisms for SnAg3.5 solder bumps on Ti/Cr-Cu/Cu and Ni(P)/Au metallization pads, Journal of Applied Physics, Vol. 96, No. 8 (2004), pp A.T.Huang, A.M.Gusak, K.N.Tu, Thermomigration in SnPb composite flip chip solder joints, Applied Physics Letter, In press, (2006).. F.T.Ouyang, A.T.Huang, K.N.Tu, Thermomigration in SnPb composite flip chip solder joints and wires, Proceedings of the 56 th Electronic Components and Technology Conference, CA, May 2006, pp C.Hillman, A.Kleyner, Assessment of eutectic solder phase growth in under-the-hood power control modules, 2005 SAE World Congress, MI, April K.M.Chen, J.D.Wu, K.N.Chiang, Effects of pre-bump probing and bumping processes on eutectic solder bump electromigration, Microelectronics Reliability, In press, (2006). 13. D.Gupta, K.Vieregge, W.Gust, Interface diffusion in eutectic Pb-Sn solder, Acta Materialia, Vol. 47, No. 1 (1999), pp J.Y.Choi, S.S.Lee, Y.C.Joo, Electromigration behavior of eutectic SnPb solder, Japanese Journal of Applied Physics, Vol. 41 (2002), pp R.Agarwal, K.N.Tu, Electromigration in eutectic tinlead solder lines at the device temperature, Proceedings of the th International Symposium on Advanced Packaging Materials: Processes, Properties and Interfaces, CA, March 2005, pp K.N.Tu, Recent advances on electromigration in verylarge-scale-integration of interconnects, Journal of Applied Physics, Vol. 94, No. 9 (2003), pp H.V.Nguyena, C.Salm, Effect of thermal gradients on the electromigration lifetime in power electronics, Proceedings of the 42 nd Annual International Reliability Physics Symposium, AZ, April 2004, pp A.P.Schwarzenberger, C.A.Ross, J.E.Evetts, A.L.Greer, Electromigration in the presence of a temperature gradient: experimental study and modeling, Journal of Electronic Materials, Vol. 17, No. 5 (1988), pp Electronic Packaging Technology Conference